Cryptosporidium parvum infects AIDS patients and other immunodeficient individuals and is a frequent cause of childhood diarrhea. Studies in human volunteers and animals have revealed pronounced differences in virulence among isolates, but the genetic determinants of virulence are unknown. The recent development in our laboratory of a method to cross C. parvum lines in mice opens new possibilitiesfor studying this pathogen using genetic methods. We have found that C. parvum lines with recombinant genotypes are produced in mixed mouse infections and that recombination between different genotypes can produce highly virulent progeny lines. Since multiple recombinant lines with different virulence properties are obtained in experimental crosses, linkage analysis can be used to identify genetic determinants of virulence, without prior knowledge of virulence determinants. It is our central hypothesis that a majority of clinically or biologically relevant phenotypic properties in C. parvum are quantitative traits controlled by multiple genes which can be identified by linkage analysis. Specific Aims: 1. Establish a high-density map of polymorphic genetic markers for C. parvum type 2. The almost completed C. parvum genome sequence will facilitate the identification of polymorphic microsatellites, restriction fragment length polymorphisms and single nucleotide polymorphisms. 2. Determine the mating structure and the inheritance of an extrachromosomal element in genotypically mixed C. parvum infections. We will test the working hypothesis that in mixed C. parvum infections fertilization results from the random fusion of identical or dissimilar gametes, and that the ratio of self- to cross-mating conforms to the model of random mating. The inheritance of a viral RNA genome will be studied in the context of this specific aim. 3. Identify virulence markers in type 2 C. parvum using Quantitative Trait Locus (QTL) analysis. This aim is based on preliminary observations that recombination between C. parvum lines generates progeny with distinct capacities to kill mice. We will test the following two working hypotheses: a) Virulence is determined by multiple genetic loci. b) QTL analysis can be used to identify genomic segments associated with virulence.